Large scale river monitoring and analysis with high resolution satellite image and meteorological information

Water is an irreplaceable strategic resource for human survival and social development. Since it flows all the time and changes very fast, it is quite important to have the capability to monitor the change of water quantity and spatial distribution accurately, quickly and timely. With the fast development of remote sensing satellite industry in China recent years, monitoring surface water on a large scale several times a year has become a task of challenge but achievable. Anhui Province is about 0.14 million km² large and full of water resources. There are 31 big rivers, grade one to grade three, flowing through Anhui Province. This paper develops the DeepLab V3+ model with river buffer strategy to carry out river extraction both in the dry and wet seasons in 2022 and uses the area accuracy, the percentage of the area of extracted water to the actual water body area, to measure river extraction accuracy. The overall area accuracy of river extraction in Anhui Province is better than 95%. 2022 satellite river monitoring result in Anhui Province shows an unique phenomena which is the total river area during the wet season does not change much compared with that during the dry season. In 2022, the total area of the 31 grade one to grade three rivers is 1375.89 km² during the dry season while 1397.78 km² during the wet season, an increase of only 21.89 km² or 1.59%. Analysis is carried out with satellite images and meteorological information, this phenomena is caused by frequent rain and snowfall in the first half of 2022 and persistent drought in the latter half of 2022. This research will be helpful for water resources monitoring, water environment and ecology protection and management in Anhui Province, China.